Q.17What special considerations apply to the management of diabetes during pregnancy?

Bibliographic InformationStudy Type & Evidence LevelNumber of PatientsPatient CharacteristicsIntervention & ComparisonFollow-up & Outcome MeasuresEffect SizeStudy SummaryReviewer Comments
Laatikainen L;Teramo K;Hieta-Heikurainen H;

1987

226
Study Type: RCT

Evidence level: 1+
13 women using continuous subcutaneous insulin infusion (CSII) 18 using conventional insulin treatment (CT)Pregnant women with Type 1 diabetes.

Country: Helsinki
Intervention: CSII

Comparison: CT
Follow-up period:

Outcome Measures:
Retinopathy
9/22 patents randomised to CSII group declined pump treatment.

Mean HbA1c decreased (P < 0.001) in both groups with no difference between groups.

Daily dose of insulin increased by 13.3 UI (32%) in the CT group and 5.7 UI (13%) in the CSII group.

The difference between groups in number of patients with progressing retinopathy was not significant.

Two patients (both CSII) progressed from background retinopathy to preproliferative retinopathy and then to proliferative retinopathy, in spite of laser treatment, following fast and significant decrease in HbA1c. Both patients were hypertensive at the beginning of pregnancy.
A rapid near normalisation of glycaemic control by CSII during pregnancy can accelerate the progress of retinopathy in poorly controlled diabetic patients.
Rosenn BM;Miodovnik M;Holcberg G;Khoury JC;Siddiqi TA;

1995 Mar

68
Study Type: Cohort

Evidence level: 2++
84 womenType 1 diabetes. Recruited before 9 weeks gestation and received intensive insulin therapy throughout pregnancy. Patients monitored glucose concentrations with memory glucometers and insulin doses were adjusted weekly accordingly.

Country:
Intervention: Intensive insulin therapy throughout pregnancy

Comparison:
Follow-up period:

Outcome Measures:
Hypoglycaemia Biochemical hypoglycaemia determined from glucometer data

Symptomatic hypoglycaemic episodes recorded weekly on standardised forms

Moderate symptomatic hypoglycaemia: Any episode that could not be resolved by the patient herself and required active assistance from another person

Severe symptomatic hypoglycaemia: an episode resulting in seizures, loss of consciousness, injury, IV glucose treatment or emergency glucagon administration.
71% of patients had at least one moderate symptomatic episode and 34% had at least one severe symptomatic episode.

19 patients had a total of 54 episodes of loss of consciousness and 12 patients had a total of 15 seizures.

Most symptomatic episodes occurred during the first half of pregnancy. The peak incidence of severe or moderate symptomatic hypoglycaemia was between 10 and 15 weeks.

There was no association between hypoglycaemia during early pregnancy and adverse pregnancy outcome.

When hypoglycaemic and nonhypoglycaemic groups were compared there were no differences in any of the commonly used indices of glycaemic control, including preprandial glucose, postprandial glucose and glycohaemoglobin concentrations.
Hypoglycaemia is a frequent complication that significantly affects the majority of women who receive intensive insulin therapy during pregnancy.

This may be related to the institution of intensive therapy which diminishes counteregulaltory responses to hypoglycaemia and increases hypoglycaemic unawareness. Alternatively it may be that pregnancy independently increases the risk of hypoglycaemia, possibly due to hormonal changes, diminished counterregulatory responses to hypoglycaemia or the nausea and vomiting of pregnancy.

At present, a measure of caution should be exercised for the occasional patient who demonatrates a tendency to have recurrent episodes of severe hypoglycaemia that cannot be resolved with the usual tactics of modifying insulin regimens, caloric intake, and physical activity.
Lapolla A;Dalfra MG;Masin M;Bruttomesso D;Piva I;Crepaldi C;Tortul C;Dalla BB;Fedele D;

2003

228
Study Type: Cohort

Evidence level: 2+
68 women treated with conventional insulin therapy 25 women treated with continuous subcutaneous insulin infusion. Pump therapy suggested because of difficulties with metabolic control.Type 1 diabetes 20/25 patients on pumps were switched from conventional therapy 1 year becoming pregnant. 21 used regular insulin and 4 used lispro. Pump therapy suggested because of difficulties with metabolic control. Pump group had greater severity of diabetes.

Country:
Intervention: Insulin pump

Comparison: conventional therapy
Follow-up period:

Outcome Measures:
Fasting blood glucose 1 hour post prandial HbA1c Spontaneous or induced abortion
Time and mode of delivery
Maternal outcomes:
Pregnancy induced hypertension
Preeclampsia
Placental insufficiency
Hydramnios
Hypoglycaemic coma
Ketoacidosis
Fetal outcomes:
Weight
Hypoglycaemia
Hypocalcaemia
Hyperbilirubinemia
Fetal distress
Asphyxia
Hyaline membrane disease
Polycythemia
Shoulder dystocia
Malformations
There was no difference in outcomes between the two groups.Insulin pump therapy is useful in problematic, complicated cases.Groups differed with regard to White class and severity of retinopathy and neuropathy.
Burkart W;Hanker JP;Schneider HP;

1988

227
Study Type: RCT

Evidence level: 1++
48 women using continuous subcutaeous insulin infusion (CSII) 41 women using intensified conventional treatment (ICT) 28 women with conventional diabetes treatment (CT)Pregnant women with Type 1 diabetes.

Country: Germany
Intervention: CSII

Comparison: ICT
CT
Follow-up period:

Outcome Measures:
HbA1c
Mode of delivery
Birthweight
>95th percentile
<10th percentile
Gestational age at delivery
Perinatal mortality
Malformations
There was no difference between CSII and ICT groups in the rate of pregnancy complications or fetal outcome. Both CSII and ICT groups had significantly better outcomes when compared with CT group.

>95th percentile:
CSII: 5/48
ICT: 6/41
CT: 8/28, P < 0.05

<10th percentile
CSII: 3/48
ICT:1/41
CT:5/28, P < 0.05

Healthy newborns:
CSII: 40/48
ICT: 34/41
CT:5/28, P < 0.05

Minor symptoms:
CSII: 8/48
ICT: 2/41
CT: 5/28

Major symptoms
CSII: 0
ICT: 2/41
CT: 22/28, P > 0.05
CSII is indicated if intensified conventional treatment does not lead to normoglycaemia
Carta Q;Meriggi E;Trossarelli GF;Catella G;Dal M;Menato G;Gagliardi L;Massobrio M;Vitelli A;

1986 Jun

432
Study Type: RCT

Evidence level: 1++
8 women with T1 diabetes in pump group
7 women with T1 diabetes in conventional treatment group
6 women with T2 diabetes in pump group
8 women with T2 diabetes in conventional treatment group 10 nondiabetic controls
15 Type 1 diabetes
14 type 2 diabetes

Country: Italy
Intervention: Continuous Subcutaneous Insulin Infusion (CSII)

Comparison: Intensive conventional treatment (ICT)
Follow-up period:

Outcome Measures:
Glycaemic control
Insulin requirement
Perinatal outcome
LGA (>90th percentile)
Type 1:
CSII: 2
ICT: 0
Type 2:
CSII: 1
ICT: 0

No significant differences in any outcome measures
CSII is a practical, safe and effective method of maintaining glycaemic control
Coustan DR;Reece EA;Sherwin RS;Rudolf MC;Bates SE;Sockin SM;Holford T;Tamborlane WV;

1986 Feb

225
Study Type: RCT

Evidence level: 1++
11 women randomised to intensive conventional insulin therapy. 11 women randomised to insulin pump therapyPregnant women with type 1 diabetes

Country:
Intervention: Insulin pump therapy (Autosyringe AS-2C, AS-6C, Lilly CPI-9100).

Comparison: Intensive conventional therapy
Follow-up period:

Outcome Measures:
Maternal hypoglycaemia:
Mild (responded to snack)
Moderate (required another person’s assistance)
Severe required glucagon or IV glucose, usually in an emergency room setting).
Mean blood glucose
HbA1c
Hyperglycaemia (>6.6)
Both treatments were associated with a significant lowering of mean blood glucose levels (P < 0.05 vs prerandomisation). Values of HbA1c fell to within the normal range in all patients.

Insulin requirements did not differ significantly between the two groups before treatment assignment. During the second and third trimesters, however, pump treated subjects required significantly less insulin than did Intensive conventional treatment group (P < 0.05).

There were no other differences between groups.

There were no pump problems or episodes of ketoacidosis. One pup-treated subject experienced two days of hyperglycaemia and ketonuria during a viral infection.
There was no difference between the two groups with respect to outpatient mean glucose levels, symptomatic hypoglycaemia, glycosylated haemoglobin levels, inpatient mean blood glucose levels, glycaemic excursions, chemical hypoglycaemia or hyperglycaemia.
Gabbe SG;Holing E;Temple P;Brown ZA;

2000 Jun

230
Study Type: Cohort

Evidence level: 2++
24 women who started continuous subcutaneous insulin infusion (CSII) during pregnancy 24 pregnant women using multiple daily injections (MDI) 12 women who were already using CSII before pregnancyPregnant women with type 1 diabetes. CSII used in women who did not have acceptable control using MDI or who requested a pump.

Country:
Intervention: CSII initiated during pregnancy

Comparison: MDI CSII initiated before pregnancy
Follow-up period:

Outcome Measures:
Mean blood glucose (from SBGM)
HbA1c
Diabetic ketoacidosis (DKA)
Severe hypoglycaemia (requiring assistance of another person and treatment with glucagon).
Caesarean delivery
Gestational age at delivery
Birth weight
LGA
SGA
Neonatal hypoglycaemia
Respiratory distress syndrome
Malformations
Perinatal death
CSII was initiated at a mean of 16.8 weeks gestation.

HbA1c declined in all study groups in the 2nd and 3rd trimesters. There was no difference in maternal or perinatal outcomes.

Before starting CSII, 8/24 women had 17 episodes of severe hypoglycaemia. After starting CSII one patient had 2 episodes (P < 0.01).

6/24 women on MDI had 8 episodes of severe hypoglycaemia.

There were no episodes of severe hypoglycaemia in 12 women who had started CSII before pregnancy.

2 patients who began CSII during pregnancy required hospitalisation for mild DKA. Both cases were attributed to localised inflammation or fibrosis at the tip of the infusion catheter. 12/24 women had pump-related problems (pump malfunction or site infection) 5/12 women who started CSII before pregnancy had pump-related problems.

Follow-up data collected for 19/24 women who started CSII in pregnancy and 10/12 women who started CSII before pregnancy.
18/19 women who had started CSII during pregnancy had continued to use CSII after pregnancy. 9/10 women who had started CSII before pregnancy had continued using pump therapy.

HbA1c following pregnancy was significantly lower in women using CSII than in women using MDI (P = 0.02).
CSII was initiated during pregnancy without deterioration in glycaemic control and was associated with maternal and perinatal outcomes comparable to those among women who were already using the pump before pregnancy or who received MDI.
Nosari I;Maglio ML;Lepore G;Cortinovis F;Pagani G;

1993

433
Study Type: RCT

Evidence level: 1++
16 women using continuous subcutaneous insulin infusion (CSII)
16 women using intensive conventional insulin therapy (ICT)
pregnant women with Type 1 diabetes

Country: Italy
Intervention: CSII

Comparison: ICT
Follow-up period:

Outcome Measures:
24 hour mean BG under a day hospital regimen
Mean home monitored BG

Frequency of home-monitored plasma values outside the target range (3.3 mmol/l-6.6 mmol/l) during pregnancy (%)HbA1c
Ketoacidosis (blood glucose >16.5 mmol/l, high positivity of urinary ketone reagent strips with arterial pH <7.2).
Severe hypoglycaemia (coma or seizure or a situation requiring hospitalisation or IV glucose or glucagons).
Progression of retinopathy
Microalbuminuria
Macro-proteinuria
Gestational age at delivery
Caesarean section
Perinatal mortality
Neonatal hypoglycaemia
Respiratory distress syndrome
Congenital malformations
Macrosomia
LGA
SGA
Good glycaemic control was achieved in both groups in the 1st trimester and maintained in the 2nd and 3rd trimester. There were no significant differences in BG levels.

The mean daily insulin requirements increased similarly in both groups.

Ketoacidosis:
CSII: 3
ICT: 0, NS

The 3 cases of ketoacidosis were due to catheter occlusion (one case) catheter leakage (one case) and pump failure (one case).

Severe hypoglycaemia:
CSII: 3
ICT: 1, NS

Intrauterine death:
CSII: 2
ICT: 1, NS

No differences in any other outcome
CSII and ICT are equally effective strategies for the treatment of pregnant diabetic women.
Simmons D;Thompson CF;Conroy C;Scott DJ;

2001 Dec

229
Study Type: Case- control

Evidence level: 2+
30 women using insulin pump 60 using conventional therapyPolynesian women with insulin-requiring GDM/type 2 with persistent hyperglycaemia despite miltiple injections of subcutanoeus insulin. Controls matched for ethnicity, type of diabetes and peak insulin dosage.

Country: New Zealand
Intervention: Insulin pumps (Nipro)

Comparison: Women treated with tablets or multiple daily injections
Follow-up period:

Outcome Measures:
2 hour postprandial glucose
Neonatal birthweight
Neonatal hypoglycaemia
Caesarean section
Women in the pump group had a higher peak inuslin dosage than women in the non pump group.

Mean 2 hour post prandial glucose (mmol/l)
Pump user: 6.6±1.8
Non pump user: 6.4±1.6
P=0.550

There was no difference in perinatal outcomes

Women using pumps put on more weight from referral (10.6±8.8 kg vs 5.0±4.9 kg, P < 0.0001) and were more likely to have baby admitted to special care baby unit (56.3%vs25.0%, P = 0.033).
Perinatal outcomes were comparable to those in women with less hyperglycaemia and lower insulin requirements

Insulin pump therapy seems to be safe and effective for maintaining glycaemic control in pregnancies complicated by GDM/type 2 diabetes and requiring large doses of insulin.
Bibliographic InformationStudy Type & Evidence LevelAim of StudyNumber of Patients & Patient CharacteristicsPopulation CharacteristicsOutcome measuresResults & CommentsStudy SummaryReviewer Comment
Rodgers BD;Rodgers DE;

1991

219
Study Type:
OtherCase series

Evidence Level: 3
Intervention:

Comparison:
37 admissions for DKA during pregnancy

Country:
Emesis and the use of β-sympathomimetic drugs were considered etiologic in 57% of cases. Patient noncompliance and physician management errors were considered etiologic in 24% and contributory in 16%.

Common physician management errors included the use of urine instead of blood to monitor maternal glucose control, failure to adhere to pregnancy standards of glucose control and failure to employ home blood glucose monitoring.

30% of patients admitted with emesis had had a prepregnancy history of diabetic gastroenteropathy, thus identifying that group at particularly high risk for DKA.
Zarkovic M;Nesovic M;Marisavljevic D;Ciric J;Stojanovic M;

1995

216
Study Type:
OtherCase study

Evidence Level: 3
Intervention: Parenteral nutrition (PN)

Glucose limited. Basal and additional insulin requirements independently assessed.

Comparison:
A pregnant woman with type 1 diabetes and hyperemesis gravidarum

Country:
Vomiting was significantly reduced after the PN started but it stopped completely after 15 days. Three days later the patient started taking small quantities of food and the PN was terminated after 23 days. There was no fetal morbidiThis case demonstrates that optimal diabetes control during PN is feasible in pregnancy. Previous experience of a medical team in treating pregnant diabetic patients is of great value in assessing daily changes in insulin requirements
Carroll MA;Yeomans ER;

2005

218
Study Type:
OtherCase study

Evidence Level: 3
Intervention:

Comparison:
A 23 year old woman of 31 weeks gestation not known to be diabetic.

Country:
On presentation heart rate was 128 beats per min. Serum blood glucose was 41.8 mmol/l. urine and serum ketones were negative. The patient was admitted with a diagnosis of pyelonephritis and hyperglycemia.

Five hours after admission IV fluid and insulin were started.

Ten hours after admission fetal death was confirmed.
Leveno KJ;Fortunato SJ;Raskin P;Williams ML;Whalley PJ;

1988 Apr 6

{Leveno, 1988 33988 /id}
Study Type:
OtherCase series

Evidence Level: 3
Intervention: Insulin pump (autosyringe AS-2C and AS-6C

Comparison: Conventional insulin therapy
28 women offered pump therapyWomen who used pumps were generally from the private sector and in better glucose control at entry.

Country:
Fasting preprandial or postprandial plasma glucose measurements perfomred during clinics.
Macrosomia
Neonatal hypoglycaemia
Respiratory distress syndrome
15/28 women offered pump therapy accepted. 5/15 subsequently discontinued use. 1 woman developed ketoacidsosis after abruptly discontinuing pump use.

Gestational age at delivery was significantly less in women refusing pump therapy (35.9 weeks vs 37.5 weeks, P < 0.05). Three malformations (2 ventricular septal defects and 1 clubfoot) in women who declined pump therapy. No malformations in women who accepted pump therapy.

No other differences between groups.
Outdated technology
Brimacombe J;

1995 Apr

{Brimacombe, 1995 34000 /id}
Study Type:
OtherCase report

Evidence Level: 3
Intervention: Parenteral nutrition

Midazolam Prochlorperazine

Comparison:
A woman with diabetes and Hyperemesis Gravidarum presenting at 17 weeks gestation

Country:
Prochlorperazine was discontinued on day three when the patient developed mild extrapyradimal symptoms Parenteral nutrition was ceased on day 6 and the midazolam on day 7. By day 8 the patient was o a full diet with slight nausea. The patient discharged herself on day 10. At 28 weeks the mother was readmitted with severe recurrent symptoms and requested C-section. TPN and pharmacotherapy reinstated (although ondanstron not midazolam due to different clinician). After 10 days woman became febrile and developed peripheral oedema.
Normal 1.5 kg baby delivered.
The combination of midazolam, prochlorperazine and TPN resulted in an improvement in symptoms which were sustained for some weeks after therapy ceased.
Bibliographic InformationStudy Type & Evidence LevelNumber of PatientsPatient CharacteristicsIntervention & ComparisonFollow-up & Outcome MeasuresEffect SizeStudy SummaryReviewer Comments
Nachum Z;Ben Shlomo I;Weiner E;Shalev E;

1999 Nov 6

{Nachum, 1999 17336 /id}
Study Type: RCT

Evidence level: 1++
392 womenWomen with gestational or pre- existing diabetes.

Country:
Israel
Intervention:
Three doses of regular insulin before meals and an intermediate insulin dose before bedtime (four times daily regimen).

Comparison: Regular and intermediate insulin in the morning and evening (twice-daily regimen).
Follow-up period:

Outcome Measures:
  1. Mean daily insulin
  2. HbA1c levels
  3. Caesarean section rate
  4. Preterm delivery
  5. Hypoglycaemia in infants
  6. Apgar score
Glycaemic control was better with the four times daily regimen than with the twice- daily regimen: in women with gestational diabetes mean blood glucose concentrations decreased by 0.19 mmol/l (95% CI 0.13 to 0.25), HbA1c by 0.3% (95% CI 0.2% to 0.4%), and fructosamine by 41 micromol/l (95% CI 37 to 45), and adequate glycaemic control (mean blood glucose concentration <5.8 mmol/l) was achieved in 17% more women (95% CI 8% to 26%); in women with pre-existing diabetes mean blood glucose concentration decreased by 0.44 mmol/l (95% CI 0.28 to 0.60), HbA1c by 0.5% (95% CI 0.2% to 0.8%), and fructosamine by 51 micromol/l (95% CI 45 to 57), and adequate glycaemic control was achieved in 31% more women (95% CI 15% to 47%).

Rates of maternal severe hypoglycaemic events, caesarean section, preterm birth, macrosomia, and low Apgar scores were similar in both groups.

In women with gestational diabetes the four times daily regimen resulted in a lower rate of overall neonatal morbidity than the twice-daily regimen (RR 0.59, 95% CI 0.38 to 0.92), and the RRs for hyperbilirubinaemia and hypoglycaemia were lower (RR 0.51, 95% CI 0.29 to 0.91 and RR 0.12, 95% CI 0.02 to 0.97, respectively). The RR of hypoglycaemia in newborn infants of women with pregestational diabetes was 0.17 (95% CI 0.04 to 0.74).
Four times daily rather than twice-daily insulin improves glycaemic control and perinatal outcomes without increasing rates of hypoglycaemia and caesarean section in the mother.
Farrar D; Tuffnell D; West J.

2007

224
Study Type:
Systematic review - meta-analysis

Evidence level: 1+
Two studies, involving 60 women with 61 pregnancies.Pregnant women with pre-existing diabetes or gestational diabetes.

Country: Both studies carried out in Italy
Intervention: This systematic review compared continuous subcutaneous insulin infusion with multiple daily injections for the treatment of diabetes during pregnancy.

Comparison:
Follow-up period:

Outcome Measures: Perinatal mortality, fetal anomalies, hypoglycemic/hyperglycemic episodes, hospital; admission and LOS.

Additional maternal and perinatal outcomes included.
No significant differences were found for the main outcome measures.

For peri-natal outcomes no significant differences in outcomes were found.

Maternal outcomes showed no significant differences in outcomes.
Provides inconclusive evidence in either direction.Neither trial included participants with GDM.

There may be serious limitations with the designs of these studies. The first trial did not describe randomization and included 2 women who had used CSII pre-conceptually. The second trial that described randomization, recruited only women with type 1 DM, described as being highly motivated.

From: Evidence tables

Cover of Diabetes in Pregnancy
Diabetes in Pregnancy: Management of Diabetes and Its Complications from Preconception to the Postnatal Period.
NICE Clinical Guidelines, No. 63.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2008 Mar.
Copyright © 2008, National Collaborating Centre for Women’s and Children’s Health.

No part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK [www.cla.co.uk]. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.